Measuring Pixel Correlations on SLAC Data Emily Phillips Longley - - PowerPoint PPT Presentation

Measuring Pixel Correlations on SLAC Data

Emily Phillips Longley

Current implementation

• Using the CrossCorrelate function of the BfTask class
• Takes in masked objects
• Outputs the correlations

○ lsst.afw.image.MaskedImageF

• Not using the Butler

Image Pre-Processing using Eotest tools

• Median bias frame subtraction
• Overscan correction using Bela’s function, doing a row-by-row

subtraction

• Masking objects by passing

○ Mask_files = (mask_dark_pixels.fits,mask_bright_pixels.fits) ○ when creating the masked_ccd object

Correlation Functions

4

Amp 1 Amp 2

Normalized Correlation Function Normalized Correlation Function

• Doing a per-amplifier analysis.

Normalized correlations Normalized correlations Y (pixels) y (pixels) x (pixels) x (pixels)

Currently not implementing

• Image processing

○ No crosstalk correction ○ No linearity correction

• Not generating the kernel
• Not performing a correction on images

Eventually Wish to Implement

• Generating the kernel and performing the correction on images!
• Specifically interested in:

○ Applying the correction to lab spots data

• Currently there is a Spots Analysis Task being developed to be a regular

eotest test

○ https://github.com/lsst-camera-dh/eotest/tree/LSSTTD-1297 ○ Added a portion to calculate the widths of the spots using the Fe55 function ○ Taking single spots data at various flux levels to measure brighter fatter